Abstracts of Technical Articles from Bell System Sources 



Ra'erberation Time and Absorption Measurements with the High 

 Speed' Level Recorder} E. H. Bedell and K. D. Swartzel, Jr. 

 It has been common in reverberation theories to neglect the effect of 

 the stationary wave pattern in a room and to assume a logarithmic 

 decay of the sound energy. In many cases this assumption of a 

 constant decay rate is not fulfilled, and in particular it is well known 

 that the decay curves as obtained with available instruments, which 

 indicate either the pressure or the velocity at a point in the sound 

 field, show marked fluctuations in the decay rate. The rate of decay, 

 in general, varies during the decay period, from point to point in the 

 room, and may depend upon the position of the sound source, and the 

 location of absorbing materials. Very rapid fluctuations in the decay 

 rate have commonly been averaged out by the measuring apparatus 

 itself, either by making the indicating instrument sluggish in its 

 action, or by measuring the average intensity over finite time intervals 

 during the decay period. The slower, and perhaps more important, 

 deviations from linearity in the decay curves have been either reduced, 

 or averaged out, by a number of expedients. Among these are the 

 use of rotating sound reflectors, or vanes, to break up the stationary 

 wave pattern in the room; the use of frequency modulated, or warble, 

 tones in place of a constant single frequency; moving the microphone 

 to a number of positions in the room; and moving the sound source. 

 Hunt has given some quantitative data on the effect of the warble tone, 

 but similar data on other methods of smoothing out the decay curves 

 are not available. This paper presents some data on the relative value 

 of the above methods of improving the measured decay curves, and 

 on the use of a motor driven rotating switch to connect in rapid 

 sequence a number of microphones, placed in different parts of the 

 room, into the measuring apparatus, for obtaining on a single curve a 

 space average of the time decay pattern. Since many of the deviations 

 from linearity in the decay curves have a "period" of 30 db or more, 

 we should expect the accuracy of our values to be a function of the 

 range through which the decay is measured, particularly when the 

 range is not large compared to the period of the deviations. This 

 effect is discussed for three values of the decay range, 30, 60 and 

 90 db. 



' Jour. Acous. Soc. Amer., January, 1935. 



360 



